national security risks

Thus usually translates to someone paid a bribe to kick out competition.

Authorities should also “urgently” consider outlawing the publication of the “weights,” or inner workings, of powerful AI models, for example under open-source licenses, with violations possibly punishable by jail time, the report says

Fuck that, so only huge corporations can have access to it. You won’t even be able to have start ups to challenge the behemoths because this would shut down any open scientific papers explaining how AI works to get started.

If you want to make a case this technology is an existential threat equivalent to nukes and any proliferation is dangerous then treat it like nukes and nationalize it and make it so only government can produce it. At least the government is nominally subject to the people instead of a bunch of companies who will happily destroy the world if it makes them an extra buck.

We’re probably nowhere near that threat though so something like this would only serve to widen the gap between the current batch of huge AI companies and smaller scale developers and enthusiasts.

We could at the bare minimum sign the autonomous weapons treaty

The U.S. government must move “quickly and decisively” to avert substantial national security risks stemming from artificial intelligence (AI) which could, in the worst case, cause an “extinction-level threat to the human species,” says a report commissioned by the U.S. government

…

many say it will arrive in the next five years

…

In other words, DARPA or a national lab have it, and China just figured it out or they’re worried they’re close.

The AI “threat” is probably the public reaction to AI

This is exactly what the big AI companies want. This will kill open source and freely available AI and anything made with it, forcing the only available options to be proprietary and lucrative.